The Agent Skills Directory

[PROMPT_INJECTION]: The SKILL.md file contains 'MANDATORY USAGE REQUIREMENT' and 'CRITICAL' instructions that attempt to override the AI agent's default behavior by forcing automatic invocation for specific events (long-running tasks, plan completion, etc.) regardless of explicit user requests.
[COMMAND_EXECUTION]: The script scripts/send.sh executes the curl command to send notification data to the Gotify API.
[DATA_EXPOSURE]: The skill accesses the sensitive file path ~/.homelab-skills/.env to retrieve GOTIFY_URL and GOTIFY_TOKEN credentials.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input in the form of notification messages and titles.
Ingestion points: Notification message and title arguments passed to scripts/send.sh.
Boundary markers: None identified in the prompt interpolation logic.
Capability inventory: Network access via curl to a user-defined endpoint.
Sanitization: The skill uses jq to safely construct the JSON payload, which prevents shell injection but does not sanitize the natural language content for downstream systems.

gotify